Sage Journals: Discover world-class research

Abstract

Head and neck squamous cell carcinoma (HNSCC) is one of the most common cancers in the world. Resistance to cytotoxic chemotherapy is a major cause of mortality in patients with HNSCC. A small subset of cancer cells called cancer stem cells (CSCs) may be key contributors to drug resistance and tumor recurrence in HNSCC. The aim of this study was to determine whether CD133, which maintains properties of CSCs, promotes chemoresistance by arresting cell cycle transition and reducing apoptosis in HNSCC cells. CD133 overexpression was examined in KB cells, and colony forming and aldehyde dehydrogenase activity assays were performed. To investigate the role of CD133 in chemoresistance, cell death was analyzed using 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT), Diff-Quick, flow cytometry, and western blot of apoptosis-related protein expression in fluorouracil (5-FU)- or cisplatin-treated cells. In addition, microarray and related protein expression assessments were performed to investigate the mechanism of chemoresistance against 5-FU and cisplatin in KB cells. Moreover, chemoresistance against 5-FU or cisplatin in a KB-inoculated mouse model was analyzed by hematoxylin and eosin staining, immunohistochemical study of CD133, and immunofluorescence of tumor tissue. In this study, we demonstrate that ectopic overexpression of CD133 significantly promotes properties of stemness in KB cell lines. Furthermore, CD133 promotes chemoresistance by arresting transition of the cell cycle and reducing apoptosis, which results in inhibition of tumor growth in 5-FU- or cisplatin-injected mouse tumor model. Taken together, our findings show that elevated levels of CD133 lead to HNSCC chemoresistance through increased stemness and cell cycle arrest.

Keywords

CD133 cancer stem cell chemoresistance 5-FU cisplatin

Introduction

Head and neck squamous cell carcinoma (HNSCC) develops in the upper aerodigestive tract, including the oral or nasal cavity, the pharynx, and the larynx. HNSCC is the sixth most common cancer in the world and remains a major health problem globally with an estimated 500,000 new cases diagnosed yearly.^1,2 Although effective therapeutic modalities such as surgery, radiation, chemotherapy, and therapeutic combinations of these approaches are used in the management of this disease, recurrent HNSCC is almost inevitable and becomes resistant to further therapies. Systemic chemotherapy, which was considered an adjuvant in the past, is currently the major approach for organ preservation strategies for HNSCC patients; it is often combined with definitive radiotherapy or induction treatments, and is no longer reserved for palliative purposes.³

Recently, interest in the cancer stem cell (CSC) model has dramatically altered the current research focus on cancer treatment and drug development.⁴ CSCs share many of the properties of their non-neoplastic counterparts.^5–7 In addition, CSCs are characterized by remarkable abilities such as extensive proliferation, self-renewal, invasion, development of distant metastasis, and drug resistance.^8,9 To investigate CSC properties, such as tumor initiation and recurrence, many studies have examined methods for the isolation of CSCs from cancer specimens or cell lines, and identified a side population of cells and several protein markers specific to CSCs.^10,11 These CSC markers include CD44, CD133, CD90, CD13, aldehyde dehydrogenase (ALDH), oval cell marker OV6, and epithelial cell adhesion molecule.^11–13 Classic chemotherapeutic agents are postulated to target differentiated cells, while CSCs appear immune to their toxicity. This research supports the premise of a significant overlap between signaling pathways involved in drug resistance and the self-renewal of cancer cells. However, the molecular mechanisms that underlie this phenomenon remain poorly explored.

CD133, which is known as prominin-1 (PROM1) and as AC133, was first described as a cell surface marker on early progenitor cells and hematopoietic stem cells. It has been used as a marker to identify CSCs derived from primary solid tumors.¹⁴ CD133 is a five-domain transmembrane protein, composed of an N-terminal extracellular tail, two small cytoplasmic loops, two large extracellular loops containing seven potential glycosylation sites, and a short C-terminal intracellular tail that can be alternatively spliced or phosphorylated.^15,16 Furthermore, CD133 expression is used as a prognostic marker of survival in squamous cell carcinoma (SCC)^17,18 and is negatively correlated with survival prognosis of patients with HNSCC. In addition, AC133, which is a glycosylated epitope of CD133 initially associated with embryonic stem cells and a variety of somatic stem cells, was extensively described as a putative CSC marker in blood, brain, colon, prostate, lung, breast, liver, and skin cancers.^19–21 Other studies found that CD133 is linked to cell metabolism as a glucose-responsive gene in myotubes²² as well as has a role in bioenergetic stress²³ and non-exposure to high oxygen tension in gliomas. However, the physiological role of CD133 in HNSCC chemoresistance has yet to be explored.

In this study, we investigated the role of CD133 in the induction of chemoresistance and CSC-like characteristics in HNSCC. To investigate this relationship, we used a KB HNSCC cell line ectopically expressing CD133 and treated it with the anti-cancer drugs, 5 fluorouracil (5-FU) or cisplatin. Our findings support the hypothesis that CD133 plays a pivotal role in chemoresistance in HNSCC by positively regulating expression of CSC-related genes, and provides a plausible mechanism for the development of chemoresistance in HNSCC.

Materials and methods

Cell culture and chemical reagents

All reagents were purchased from Sigma–Aldrich (St. Louis, MO, USA). KB cells were purchased from the Korean Cell Line Bank (KCLB no. 10017, Seoul, Korea) and maintained in RPMI 1640 medium (Welgene, Daegu, Korea) supplemented with 10% heat-inactivated fetal bovine serum (Gibco BRL Life Technologies, Rockville, MD, USA) and a 10% antibiotic–antimycotic solution (Welgene) at 37°C in a 5% CO₂ humidified chamber. After seeding, used medium was replaced with fresh medium and adherent cells were allowed to reach approximately 70% confluence. Cells were then detached using trypsin–ethylenediaminetetraacetic acid (EDTA) (Gibco BRL) and re-plated (subcultured) in 6-well plates for each experiment.

Cloning of human CD133

HT29 colon cancer cells (KCLB no. 30038) were used as a source of CD133 protein. Cloning of CD133 was performed as described previously.²⁴ 293T cells (KCLB, no. 21573) were used as a positive control to confirm the translated fusion product. Briefly, 293T cells were grown to near-confluence in a 60-mm dish, followed by transient transfection with either pcDNA3.1/NT-GFP or pcDNA3.1/NT-GFPCD133 using 0.6 µL of the FuGENE HD transfection reagent (Promega, Madison, WI, USA) with 0.2 µg of plasmid. After 48 h, cells were gently washed with ice-cold phosphate-buffered saline (PBS; Welgene).

Establishment of stably transfected KB-GFP control cells and KB-GFPCD133 cells expressing the fusion protein

Stable cell lines overexpressing green fluorescent protein (GFP) or the CD133–GFP fusion protein were generated by transfecting confluent KB cells in 100-mm plates with 20 µg of pcDNA3.1/NT-GFPCD133 or pcDNA3.1/NT-GFP as described previously.²⁴

Western blot

Once cell cultures were approximately 70% confluent, the medium was removed, cells were washed twice with PBS, and cell lysates were prepared in 200 µL of cold lysis buffer (1% NP-40, 50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 0.02% sodium azide, 150 mg/mL phenylmethylsulfonyl fluoride (PMSF), 2 mg/mL aprotinin, 20 mg/mL leupeptin, and 1 mg/mL pepstatin A). Approximately 30 µg of protein from each sample was separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) on a 10% gel and transferred to a polyvinylidene difluoride membrane (Amersham, Piscataway, NJ, USA). The membrane was incubated for 0.5 h with a blocking solution consisting of 5% skim milk in Tris-buffered saline and Tween 20 (TBST: 2.42 g/L Tris-HCl, 8 g/L NaCl, 0.1% Tween 20, pH 7.6), rinsed briefly in TBST, and incubated overnight at 4°C with the appropriate primary antibody: anti-CD133 (1:500; MyBioSource, San Diego, CA, USA), anti-OCT4 (1:1000; Cell Signaling Technology, Danvers, MA, USA), anti-NANOG (1:1000; Cell Signaling Technology), anti-procaspase 3 (1:1000; Santa Cruz Biotechnology, Santa Cruz, CA, USA), anti-procaspase 9 (1:1000; Santa Cruz Biotechnology), anti-BAX (1:1000; Santa Cruz Biotechnology), anti-BCL2 (1:1000; Santa Cruz Biotechnology), anti-cyclin E (1:1000; Santa Cruz Biotechnology), anti-cyclin A (1:1000; Santa Cruz Biotechnology), or anti-cyclin D (1:1000; Santa Cruz Biotechnology). Mouse monoclonal anti-glyceraldehyde 3-phosphate dehydrogenase (GAPDH) IgG (1:2500; Santa Cruz Biotechnology) was used as a control. After rinsing with TBST, the membrane was incubated for 1 h with an anti-rabbit or anti-mouse horseradish peroxidase–conjugated secondary antibody (1:2000; Santa Cruz Biotechnology). Finally, the membrane was washed in TBST, and protein immunoreactivity was detected using an enhanced chemiluminescence detection kit (Amersham). Protein levels were determined by densitometry analysis using the ImageJ software (rsb.info.nih.gov/ij).

Confocal microscopic analyses

Cells were counterstained using 4′,6-diamidino-2-phenylindole (DAPI) in ProLong Gold antifade mounting medium (Invitrogen, Carlsbad, CA, USA) to visualize nuclear morphology. Digital images were captured using a TCS SP5 AOBS laser-scanning confocal microscope (Leica Microsystems, Heidelberg, Germany) with a 20× objective at the Korea Basic Science Institute Gwangju Center at Gwangju, Korea.

Colony forming assay

Non-adherent 24-well culture plates were coated with a 10% polyHEMA (Sigma–Aldrich) solution in absolute ethanol and dried overnight. After seeding, cells were incubated in serum-free Dulbecco’s modified Eagle’s Medium (DMEM; Welgene) supplemented with 200 ng/mL epithelial cell growth factor (R&D systems, Minneapolis, MN, USA), 20 ng/mL basic fibroblast growth factor (Sigma–Aldrich), and B-27 supplement (Invitrogen). After an incubation of 5 days, the number of spheroids in each well was counted under a light microscope (Zeiss, Zena, Germany).

Flow cytometry assessment of ALDH activity

Flow cytometry was used to examine tumor-derived cells using the ALDEFLUOR kit (StemCell Technologies, Vancouver, Canada). Cells were suspended in ALDEFLUOR assay buffer containing ALDH1 substrate (BODIPY-aminoacetaldehyde; 1 M per 10⁶ cells) and incubated for 30 min at 37°C. Some cell batches were incubated with the specific ALDH1 inhibitor diethylaminobenzaldehyde (DEAB; 50 mM) and were used as negative controls. Cells were washed twice with washing buffer and analyzed by flow cytometry (FC-500, Beckman Coulter, Carlsbad, CA, USA). Approximately 10,000 counts were accumulated for each sample.

Cell viability assay

Cell viability was evaluated using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT; Sigma–Aldrich) assay. Cells were seeded at 5 × 10³ in 200 µL of medium in 96-well plates and cultured for 1 day at 37°C. After each experiment, medium was removed and 20 µL MTT (50 µg/mL) was added to each well. Cells were incubated at 37°C for 4.5 h to allow color formation, and the formazan product was solubilized by addition of 50 µL dimethyl sulfoxide (Calbiochem Bio-Mol, La Jolla, CA, USA). Optical density at 570 nm was measured using a microtiter plate reader (Bio-Rad, Hercules, CA, USA).

Histomorphological evaluation (Diff-Quick staining)

To analyze morphological changes, cells were stained using the Diff-Quick method (International Reagents Corp., Kobe, Japan). Diff-Quick staining was originally designed to incorporate cytoplasmic staining (pink) with nuclear staining (blue) for rapid cytological detection. Cells (1 × 10⁶) were seeded on coverslips with 2 mL of medium in 6-well plates. Following 24 h incubation, the coverslips were allowed to air dry for 10 min. Next, the coverslips were treated with Diff-Quick fixative for 60 s, Diff-Quick Solution I for 60 s, and Diff-Quick Solution II for 60 s, washed in deionized water for 10 s, allowed to air dry, mounted on slides, and examined by light microscopy.

Flow cytometry analysis of apoptosis

Cells in early and late apoptotic stages were quantified using an Annexin V-FITC/PI double staining assay. Cells were harvested with trypsinization and washed twice with PBS. Cells were resuspended in 500 µL binding buffer, followed by staining with 10 µL Annexin V and 5 µL propidium iodide (PI) in the dark at room temperature (RT) for 15 min. Stained cells were immediately examined using a FACS flow cytometry analyzer (Beckman Coulter) with wavelength emission filters of 488–530 nm for the green fluorescence of Annexin V (FL1) and of 488–630 nm for the red fluorescence of PI (FL2). At least 10,000 cells per sample were acquired to ensure sufficient data collection.

Gene expression analysis using microarrays

For RNAs of KB^Vec or KB^CD133+, the synthesis of target complementary RNA (cRNA) probes and hybridization were performed using Agilent’s LowInput QuickAmp Labeling Kit (Agilent Technology, Santa Clara, CA, USA) according to the manufacturer’s instructions. Briefly, 25 ng of both total RNA and T7 promoter primer were mixed and incubated at 65°C for 10 min. The cDNA master mix (5X First strand buffer, 0.1M DTT, 10 mM dNTP mix, RNase-Out, and MMLV-RT) was prepared and added to the reaction. Samples were incubated at 40°C for 2 h followed by reverse transcription polymerase chain reaction (RT-PCR), and double stranded DNA (dsDNA) synthesis was terminated by incubating at 70 C for 10 min. The transcription master mix was prepared using the manufacturer’s protocol (4X transcription buffer, 0.1M DTT, dNTP mix, 50% polyethylene glycol (PEG), RNase-Out, Inorganic pyrophosphatase, T7-RNA polymerase, and Cyanine 3/5-CTP). Transcription of dsDNA was performed by adding the transcription master mix to the dsDNA reaction samples and incubating at 40°C for 2 h. Amplified and labeled cRNA was purified using RNase mini columns (Qiagen, Alameda, CA, USA) according to the manufacturer’s protocol. The labeled cRNA target was quantified using a ND-1000 spectrophotometer (NanoDrop Technologies, Inc., Wilmington, DE, USA). After checking labeling efficiency, 850 ng of both cyanine 3-labeled and cyanine 5-labeled cRNA targets were mixed, and cRNA fragmentation was performed by adding 10X blocking agent and 25X fragmentation buffer, followed by incubation at 60 C for 30 min. The fragmented cRNA was resuspended using 2X hybridization buffer and directly pipetted onto an assembled Agilent Human GE 4X 44K v2 Microarray. The arrays were hybridized at 65°C for 17 h using an Agilent Hybridization oven (Agilent Technology). Hybridized microarrays were washed using the manufacturer’s washing protocol (Agilent Technology).

Data acquisition and analysis

The hybridization images were scanned using an Agilent DNA microarray Scanner (Agilent Technology), and data quantification was performed using Agilent Feature Extraction software 10.7 (Agilent Technology). The average fluorescence intensity for each probe was calculated, and local background was subtracted. All data normalization and identification of fold-changed genes were performed using GeneSpringGX 7.3.1 (Agilent Technology). Genes were filtered with removing flag-out genes in each experiment. Intensity-dependent normalization (locally weighted scatterplot smoothing (LOWESS)) was performed, where the ratio was reduced to the residual of the LOWESS fit of the intensity versus the ratio curve. The averages of normalized ratios were calculated by dividing the average of the normalized signal channel intensity by the average of the normalized control channel intensity. Functional annotation of genes was performed according to the Gene Ontology^TM Consortium (www.geneontology.org/index.shtml) using GeneSpring GX 7.3.1.

Flow cytometry analysis of cell cycle

Cells were trypsinized to obtain a single-cell suspension, harvested by centrifugation, and washed with PBS. After fixation in ice-cold 70% ethanol at 4°C overnight, cells were collected and washed twice with PBS. The cells were then stained with PI (3 µg/mL) for 15 min. Afterwards, DNA from 10⁴ of G0/S phase cells was quantified using a flow cytometer (Beckman Coulter).

Animal experiments

Five week-old male athymic nude mice (BL-6/nu, Orient Bio Co. LTD, Seoul, Korea) were housed under controlled light conditions and fed ad libitum. Xenograft tumors were generated by injecting KB^Vec or KB^CD133+ into the flanks of mice and treated with either saline, 5-FU, or cisplatin.²⁵ The animal protocol was approved by the Institutional Animal Care and Use Committee, Chosun University, Gwangju, Korea (CIACUC2016-S0007).

Following treatment, mice were sacrificed by intraperitoneal administration of 0.4% sodium pentobarbital (1 mL/kg). All tumor samples were dissected and tumor weights were measured. Afterwards, all tumor samples were fixed in 10% buffered formalin (Merck, Darmstadt, Germany), embedded in paraffin, sectioned, and mounted without staining for monitoring of GFP immunofluorescence or stained with hematoxylin and eosin (H&E; Merck) for histological analysis.

Immunohistochemical study

Three micrometer-thick sections were deparaffinized in three changes of xylene and rehydrated in a graded series of ethanol to distilled water. For antigen retrieval, slides were placed in 0.01 M citrate-buffer pH 6.0 and heated in a steamer for 30 min. Endogenous peroxidases quenching was done by the slides incubating in 3% hydrogen peroxide for 20 min at RT. Sections were incubated overnight at 4°C with a 1:50 dilution primary antibody: mouse polyclonal anti-CD133 (Mybiosource, San Diego, CA). Subsequently, sections were incubated with biotinylated secondary antibody (LSAB; Dako Cytomation, Glostrup, Denmark) for 30 min, washed in PBS, and incubated with streptavidin–peroxidase conjugate (LSAB, Dako Cytomation) for 30 min. The reaction was developed using 3,30-diaminobenzidine tetrahydrocloride (Sigma–Aldrich) for 5 min. Slides were briefly counterstained in hematoxylin, dehydrated, and cover slipped. Negative and positive controls were run simultaneously. Positive controls were represented by mammary tissue.

Statistical analysis

All analyses were performed using GraphPad Prism Ver. 6.0 (GraphPad Software Inc. San Diego, CA, USA). All data were expressed as mean ± standard error of the mean (SEM) with statistical significance considered with p < 0.05. Statistical comparisons between control and treatment groups were determined using an unpaired t test or a one-way analysis of variance (ANOVA) with Tukey’s post-hoc test.

Results

Stable overexpression of CD133 in KB cells

To determine the effect of CD133 overexpression in vitro, we established stable cell lines overexpressing CD133 under the control of a constitutive promoter. Specifically, we transfected the HNSCC KB cell line with either GFP-tagged CD133 or an empty vector simply containing GFP. A representative CD133-GFP clone was selected from KB^CD133+ and compared with a control cell line transfected with plasmid expressing GFP alone (KB^Vec). Transiently transfected 293T cells with GFP (293T^Vec) or CD133-GFP (293T^CD133+) were used as positive controls. At the protein level, basal expression of CD133 was very low in KB^Vec, but significantly higher in the stable CD133-GFP-transfected cells (KB^CD133+) and transiently transfected 293T cells (Figure 1(a)). KB^CD133+ cells expressed green fluorescence in the cytosol and cell membrane (Figure 1(b)).

Figure 1.

CD133 overexpression in KB cell lines. (a) Western blot analysis of CD133 in KB^Vec (controls) and KB^CD133+ (CD133-overexpressing) cells. 293T cells were transiently transfected with pcDNA3.1/NT-GFP or pcDNA3.1/NT-GFPCD133, and 293T^Vec/293T^CD133+ cells were used as a positive control. (b) Confocal microscopy of GFPCD133 expression in stably or transiently transfected KB-GFPCD133/293T-GFPCD133 cells. Nuclei in both sets of images were stained with DAPI (blue). Images were taken at 630× magnification. Scale bar, 10 µm.

Stem cell properties in KB cells following ectopic overexpression of CD133

To examine the role of CD133 in cancer stemness, we analyzed cells for CSC-like properties (i.e. increased ability to form tumor spheres, ALDH activity, and expression of stem cell–like markers such as ALDHA1, OCT4, and NANOG). We found that KB^CD133+ cells had increased expression of OCT4, ALDHA1, and NANOG, consistent with increased stemness (Figure 2(a)). In addition, stemness was confirmed using the colony formation assay. KB^CD133+ cells were significantly better at forming colonies than that found in control cells (Figure 2(b)).

Figure 2.

Effect of CD133 overexpression on markers associated with stemness. (a) Expression of stemness-related proteins was characterized by Western blot. GAPDH was used as a loading control. Results are representative of three separate experiments with comparable results. (b) Stable overexpression of CD133 led to a significant increase in the ability of KB cells to form colonies. Images were taken at 100× magnification. Scale bar, 100 µm. Data in the bar graphs are expressed as the mean ± standard deviation (SD), *, p < 0.05. (c) Representative flow cytometry plots showing aldehyde dehydrogenase (ALDH) activity in prostate cancer cells. In each case (Vec or CD133+ cells), the plots on the left show negative controls cells treated with the ALDH inhibitor, diethylaminobenzaldehyde (DEAB). The gated cell populations (labeled B) represent the ALDH-positive subpopulations.

Next, using the ALDEFLUOR assay, we assessed ALDH activity in KB^CD133+ cells (Figure 2(c)), and found that 1.84% of KB^Vec cells were ALDEFLUOR-positive. In contrast, we found that 5.3% of KB^CD133+ cells were ALDEFLUOR-positive. These results demonstrate that ectopic overexpression of CD133 in KB cells results in increase of ALDH activity, colony forming ability, and expression of ALDHA1, OCT4, and NANOG.

Effects on apoptotic cell death by 5-FU or cisplatin treatment in CD133-overexpressing KB cells

To determine whether ectopic overexpression of CD133 decreased the efficacy of 5-FU or cisplatin, we measured the cytotoxicity of both drugs at various concentrations with a 24 h exposure in KB^Vec and KB^CD133+ cells. We found that cell viability of KB^CD133+ cells was higher than that found in KB^Vec cells during the entire incubation period (Figure 3(a)), with an LD₅₀ of 0.6 mM and 0.8 mM for 5-FU and cisplatin, respectively, in KB^Vec cells. These LD₅₀ concentrations were used in follow-up experiments.

Figure 3.

CD133 overexpression promotes chemoresistance in KB cells against 5-FU or cisplatin. (a) Cell viability was determined by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. KB^Vec or KB^CD133+ cells were exposed to various doses (0, 0.025, 0.05, 0.1, 0.2, 0.4, 0.6, 0.8, 1, 2, and 4 mM) of 5-FU or cisplatin for 24 h. The results are presented as the mean ± SD of three independent experiments, and significant differences (p < 0.05) compared to the untreated group are marked with an asterisk. (b) Cell viabilities were determined at 0, 3, 6, 12, and 24 h in the presence of 5-FU (0.6 mM) or cisplatin (0.8 mM). (c) Histomorphological evaluations by Diff-Quick staining at 24 h are shown. KB^Vec or KB^CD133+ cells were exposed to 5-FU (0.6 mM) or cisplatin (0.8 mM). All magnifications were at 200×. Scale bar, 20 µm. (d) Apoptosis measured by flow cytometry after double staining with Annexin-V(FITC, FL1) and propidium iodide (PI, FL2). Comparable results were obtained in three separate experiments. (e) Western blot analysis of apoptosis-related proteins in 5-FU (0.6 mM) treated KB^Vec or KB^CD133+ cells. GAPDH was used as a loading control. Results are representative of three separate experiments with comparable results. (f) Western blot analysis of apoptosis-related proteins in cisplatin (0.8 mM) treated KB^Vec or KB^CD133+ cells.

To evaluate the effects of 5-FU or cisplatin on cytotoxicity of KB^Vec and KB^CD133+ cells over time, treatment using 0.6 mM of 5-FU or 0.8 mM cisplatin for 0, 3, 6, 12, and 24 h were performed in KB^Vec and KB^CD133+ (Figure 3(b)). KB^Vec cells treated with 5-FU or cisplatin for 0, 3, 6, 12, and 24 h showed a time-dependent decrease in cell viability. Of importance, we found no significant cytotoxicity in KB^CD133+ cells using 5-FU.

The morphological features of KB^Vec and KB^CD133+ cells treated with 0.6 mM of 5-FU or 0.8 mM of cisplatin at 24 h were observed to determine whether the anti-cancer drugs induced morphological changes associated with apoptosis (Figure 3(c)). KB^Vec cells treated with either 5-FU or cisplatin showed a reduced number of cells compared to that found in KB^CD133+ cells. Furthermore, KB^CD133+ cells displayed intact cellular morphologies, but 5-FU- or cisplatin-treated KB^Vec cells showed evidence of apoptosis (i.e. apoptotic bodies with membrane blebs characteristic of apoptotic cell death).

Using flow cytometry analysis with an Annexin V-FITC/PI double staining assay, we evaluated 5-FU- or cisplatin-induced apoptotic cell death in KB^Vec and KB^CD133+ cells (Figure 3(d)), and found that the B2 region, which represents a late stage of apoptotic cells, was increased from 2.4% to 2.9% in 5-FU-induced and 17.3% in cisplatin-induced KB^Vec cells. However, we found that the B2 region in KB^CD133+ cells was lower being 1.9% in 5-FU-induced and 9.4% in cisplatin-induced KB^CD133+ cells compared to that found in KB^Vec cells.

We further characterized changes to key proteins involved in the apoptotic pathway. Using western blot analysis, we found that the levels of pro-caspase 3 to caspase 3 in KB^Vec cells were significantly increased by 5-FU or cisplatin treatment in a time-dependent manner (Figure 3(e) and (f)). However, we found there were no changes from pro-caspase 3 to caspase 3 in 5-FU treated KB^CD133+ cells. In addition, expression of the anti-apoptotic protein Bcl-2 (also known as BCL2) in KB^Vec cells was significantly inhibited by 5-FU as well as by cisplatin compared to that found in KB^CD133+ cells. Expression of another important apoptotic protein, BAX, was also significantly increased by 5-FU or cisplatin treatment in KB^Vec cells, but was decreased in KB^CD133+ cells. Our findings of reduced Bcl-2 expression in addition to increased expression of BAX and caspase 3 activation in KB^CD133+ cells by 5-FU treatment particularly indicate that CD133 overexpression drives cells to undergo apoptosis and develop chemoresistance to 5-FU.

Effects on cell cycle arrest in CD133-overexpressing KB cells

Using GeneSpringGX software, differentially expressed genes with a functional annotation were identified in both KB^Vec and KB^CD133+ cells (Figure 4(a) and (b), GSE92533). Compared to KB^Vec and KB^CD133+ cells, RNA splicing and cell cycle–related genes were significantly upregulated. Therefore, to investigate the effect on resistance against cancer drugs, cell cycle checkpoint–related genes that are differentially expressed because of CD133+ overexpression in KB cells were identified. The differentially expressed genes with a fold–change >2 in KB^CD133+ cells compared to that found in KB^Vec control cells were RPS27L, CCNE2, CCNA2, CCND1, and CENPF. Their relative mRNA expression levels were significantly downregulated in KB^CD133+ cells compared to that found in KB^Vec cells (Figure 4(c)), a finding that indicates cell cycle checkpoint–related genes were downregulated because of ectopic overexpression of CD133 in KB cells. Therefore, we investigated protein expression related to cell cycle checkpoints with 5-FU or cisplatin treatment in KB^Vec and KB^CD133+ cells (Figure 4(d)). We found that expression of p53, which has a crucial role in the cell cycle checkpoint in KB^Vec cells, was decreased compared to that found in KB^CD133+ cells. However, treatment with either 5-FU or cisplatin led to significantly increased p53 expression in KB^Vec cells. In addition, expression of cyclin E, which advances the cell cycle, was increased in KB^Vec cells compared to that found in KB^CD133+ cells, but 24 h treatment with either 5-FU or cisplatin resulted in a significant decrease in its expression in KB^Vec cells.

Figure 4.

CD133 overexpression promotes chemoresistance in KB cells against 5-FU or cisplatin by cell cycle checkpoint regulation. (a) Microarray analysis (left) and scatterplot of differentially expressed genes in the KB^Vec vs. KB^CD133+ cells (right). (b) Cluster analysis of differentially expressed mRNAs (left) and heatmap of cell cycle checkpoint gene expression (right) in KB^Vec vs. KB^CD133+ cells as determined by microarray analysis of cDNA expression. (c) mRNA expression of cell cycle checkpoint by microarray analysis. (d) Western blot analysis of cell cycle checkpoint–related protein expression in 5-FU (0.6 mM) or cisplatin (0.8 mM) treated KB^Vec or KB^CD133+ cells. (e) Cell cycle analysis of KB^Vec or KB^CD133+ cells. A means the G0 and G1 phases, C indicates the G2 and M phases, and D means the S phase. (f) Western blot analysis of stemness-related protein expression in 5-FU (0.6 mM) or cisplatin (0.8 mM) treated KB^Vec or KB^CD133+ cells for 24 h.

To determine whether the cell cycle pattern changed because of CD133 overexpression in KB cells, we performed flow cytometry analysis in KB^Vec and KB^CD133+ cells (Figure 4(e)). We found that in KB^Vec cells, the G0/G1 stage was 58.1%, S stage was 28.6%, and the G2/M stage was 14.3%. However, we found that G0/G1 was 84.8%, the S stage was 12.6%, and the G2/M stage was 0.9% in KB^CD133+ cells. Taken together, overexpression of CD133 in KB cells results in cell cycle arrest that in turn induces development of chemoresistance against 5-FU or cisplatin.

Histological findings and distribution of CD133-overexpressing KB cells

Finally, to determine whether overexpression of CD133+ increases chemoresistance in a xenograft model, treatment using 5-FU or cisplatin was performed in KB^Vec or KB^CD133+ cells inoculated in mice, and tumor growths were measured (Figure 5(a)). Treatment using either 5-FU or cisplatin significantly decreased both tumor growth and weight in KB^Vec-inoculated mice (Figure 5(b) and (c)). However, treatment using either 5-FU or cisplatin did not lead to decreased tumor growth or weight in KB^CD133+-inoculated mice. Furthermore, H&E staining showed that inoculation of KB^CD133+ cells in mice exhibited blood vessels and necrotic lesion similar to that found in KB^Vec-inoculated mice (Figure 5(d)). However, 5-FU or cisplatin treatment did not decrease tumor lesion in KB^CD133+-inoculated mice compared to that found using KB^Vec cells, a finding that indicates KB^CD133+ cells have an increased resistance against 5-FU and cisplatin. In addition, we found GFP-fluorescence and immunoexpression of CD133 in KB^CD133 cells in 5-FU- or cisplatin-treated mice compared to that seen in KB^Vec cells (Figure 5(e) and (f)). Taken together, CD133 overexpression results in decreased tumor degeneration by 5-FU or cisplatin treatment in a mouse model.

Figure 5.

Effect of 5-FU or cisplatin treatment on in vivo tumor formation ability by KB^Vec or KB^CD133+ cell inoculation. (a) Schematic diagram of the 5-FU and cisplatin treatment protocol. KB^Vec or KB^CD133+ cells were subcutaneously inoculated into immunodeficient mice to form in vivo xenografts. (b) Gross morphology of xenotransplanted tumor tissue (n = 3). (c) Tumor weight, **p < 0.01 vs. group (n = 4). (d) Hematoxylin and eosin staining of KB^Vec or KB^CD133+ tumors reveal cellular heterogeneity of the tumors; original magnification 10×. (e) Immunofluorescence of green fluorescent protein (GFP) in tumor. Images were taken at 100× magnification. Scale bar, 100 µm. (f) Immunohistochemical study of CD133 positive cells in tumor. Images were taken at 100× magnification. Scale bar, 100 µm.

Discussion

HNSCC is difficult to treat clinically and is resistant to conventional chemotherapy and radiation treatment modalities.²⁶ Although 5-year survival rates of HNSCC are high even at an advanced stage, the prognosis of HNSCC remains poor because many cases of HNSCC eventually become chemotherapy-resistant in addition to recurrence and development of metastasis. Current conventional treatments of HNSCC eliminate most cells within a tumor, but advanced cancers still progress to an incurable and metastatic disease state.²⁷ Nowadays, many studies have proposed that CSCs may be a major cause of therapy resistance in advanced HNSCC. These small populations of cells possess unlimited self-renewal capacities and can regenerate tumorigenic progenies, and therefore, have an essential role in HNSCC therapy resistance, metastasis, and disease relapse.²⁸ Therefore, targeting these CSCs may be a promising and effective strategy to combat HNSCC. To investigate the resistance of conventional anti-cancer drugs, we first sought to identify CSC-like cells by inducing ectopic overexpression of CD133 in the HNSCC KB cell line. Acquisition of the capacity for self-renewal and stemness is essential for tumorigenesis, and for continued accumulation of oncogenic changes that lead to cancer.²⁴ Overexpression of CD133 in KB cells leads to expression of several factors related to stemness, including expression of OCT4, ALDHA1, and NANOG, which provide cells with colony forming abilities and ALDH activity. Our findings confirmed that HNSCC cells had potentiated stem cell features because of CD133. Overexpression of CD133 in HNSCC cells facilitated both in vitro self-renewal of HNSCC cells and in vivo xenograft growth.

The results of the present study that demonstrate CD133 expression increases resistance of KB cells to chemotherapeutic reagents such as 5-FU and cisplatin strongly suggest the necessity for the implementation of therapeutic regimens that consider CD133 as a priority target. Especially, 5-FU treatment using higher doses over 0.6 mM could not effectively cause cell death in KB^CD133+ cells compared to that found in KB^Vec cells. In contrast to KB^CD133+ cells, treatment with 5-FU of KB^Vec cells led to the induction of apoptosis via upregulation of BAX and downregulation of Bcl-2, and finally, caspase 3 activation. Oligomerization of BAX results in the formation of pores in the outer mitochondrial membrane, which promotes the release of cytochrome c, whereas Bcl-2 binds to BAX and inhibits the formation of BAX oligomers.²⁹ Thus, increasing the level of Bcl-2 relative to BAX (i.e. a lower BAX/Bcl-2 ratio) promotes the mitochondrial retention of cytochrome c and results in the reduced apoptotic cell death observed in KB^CD133+ cells. These results indicate that CD133 overexpression in HNSCCs leads to inhibition of anti-cancer drug–induced apoptosis, especially 5-FU, resulting in the onset of chemoresistance.

Next, we asked how CD133 promotes reduction of apoptosis in anti-cancer drugs used to treat HNSCC. To answer that question, we performed microarray analysis in KB^Vec and KB^CD133+ cells. We found by microarray data analysis that many genes and pathways are involved in 5-FU or cisplatin resistance, which highlights the complexity of chemoresistance mechanisms, as demonstrated both by our findings and that of others made previously.^30–32 We found that 16% of cell cycle–related genes were upregulated because of CD133 overexpression, whereas genes associated with cell cycle checkpoints such as RPS27L, CCNE2, CCNA2, and CCND1 were significantly decreased. We also found that because of CD133 overexpression, p53 protein expression was significantly increased while cyclin E expression was decreased, which resulted in decreased S and G2/M stages. CSCs are quiescent cells, and since most conventional chemotherapeutic drugs interfere with the growth of rapidly dividing cancer cells, it is believed that CSCs are spared from these treatments. Consequently, this leads to chemoresistance, metastasis, and tumor recurrence. We reported for the first time the cell cycle status of KB^Vec and KB^CD133+ cells. Our findings clearly show that KB^CD133+ cells were quiescent compared to that found in KB^Vec cells. Furthermore, this observation is in agreement with existing stem cell theory, further demonstrating the slow-cycling nature of stem cells.

In HNSCC, cisplatin and 5-FU treatment is an accepted induction chemotherapy regimen.³³ Multiple studies have reported the anti-tumor effects of 5-FU or cisplatin in treating head and neck cancer. 5-FU was the systemic agent of choice for advanced HNSCC for many years, and several recent studies have demonstrated a valuable role for 5-FU in treatment protocols. Also, cisplatin, which is a platinum-based chemotherapeutic agent, is currently the backbone for several chemotherapy combinations used in treating metastatic and locally recurring HNSCC as well as other cancers.³⁴ In this study, we have shown that KB^Vec and KB^CD133+ cells exposed to either 5-FU or cisplatin developed resistance and acquired protein expression, including decreased p53 and increased cyclin E. Furthermore, CD133+ cells treated with 5-FU or cisplatin acquired cell cycle progression–related protein expression, which resulted in survival in 5-FU- or cisplatin-treated mice. The mechanism of 5-FU has been associated with inhibition of thymidylate synthase (TS) and the incorporation of 5-FU into RNA and DNA,³⁵ which results in the cytotoxic action. Evidence that this mechanism contributes to toxicity was obtained from studies with uridine protection. In normal epithelium of the intestine, p53-dependent apoptosis was only inhibited by uridine, and not by thymidine, after exposure to 5-FU.³⁶ In addition, the cytotoxic effects of cisplatin largely depend on the activation of apoptosis. Cisplatin exerts anti-cancer effects via multiple mechanisms, yet its most prominent (and best understood) mode of action involves the generation of DNA lesions, followed by the activation of the DNA damage response, and the induction of mitochondrial apoptosis.³⁷ Prolonged arrest of cell division eventually activates a checkpoint and induces programmed cell death. The mechanism that is relevant to cisplatin resistance involves overexpression of the copper transporter CTR1 (also known as SLC31A1), which is a major influx transporter, and resulting modulation of p53 activities.³⁸ Therefore, in the present study, CD133 overexpression led to cell cycle arrest in KB cells and induced chemoresistance to anti-cancer drugs, especially 5-FU. Although many current therapeutic strategies target a specific self-renewal pathway using a unique drug/inhibitor, such compounds are unlikely to be effective considering the involvement of multiple interactive pathways that drive self-renewal capacity in CSCs. Hence, targeting one specific pathway will allow cancer cells to “escape” via alternative pathways, and thus, such approaches will be inadequate to eliminate CSCs.

In summary, our findings show that CD133 has a functional role in regulating stem cell properties in HNSCCs by promoting colony formation, ALDH activity, and increased expression of CSC markers such as OCT4 and NANOG. In addition, CD133 overexpression leads to increased gene expression related to the cell cycle checkpoint and induced cell cycle arrest that results in induction of anti-apoptosis in 5-FU- or cisplatin-treated HNSCC cells and inhibition of tumor growth in 5-FU- or cisplatin-inoculated mouse models. Our data substantiate the role of chemoresistance and that of CSCs in HNSCC. Therefore, our results may facilitate the development of a novel classification system and therapeutic strategies for HNSCC.

Footnotes

Acknowledgements

Junyoung Lee and Mineon Park contributed equally to this work.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

This work was supported by a grant from the Clinical Medicine Research Institute of the Chosun University Hospital (2016).

References

Siegel

Ward

Brawley

. Cancer statistics, 2011: the impact of eliminating socioeconomic and racial disparities on premature cancer deaths. CA Cancer J Clin 2011; 61: 212–236.

Mery

Guy

Espenel

. Targeting head and neck tumoral stem cells: from biological aspects to therapeutic perspectives. World J Stem Cells 2016; 8: 13–21.

Lamont

Vokes

. Chemotherapy in the management of squamous-cell carcinoma of the head and neck. Lancet Oncol 2001; 2: 261–269.

Visvader

Lindeman

. Cancer stem cells in solid tumours: accumulating evidence and unresolved questions. Nat Rev Cancer 2008; 8: 755–768.

Tirino

Desiderio

d’Aquino

. Detection and characterization of cd133+ cancer stem cells in human solid tumours. PLoS One 2008; 3: e3469.

Liu

Guo

. G9a is essential for EMT-mediated metastasis and maintenance of cancer stem cell-like characters in head and neck squamous cell carcinoma. Oncotarget 2015; 6: 6887–6901.

Angelastro

Lame

. Overexpression of cd133 promotes drug resistance in c6 glioma cells. Mol Cancer Res 2010; 8: 1105–1115.

Fessler

Dijkgraaf

De Sousa

EMF

. Cancer stem cell dynamics in tumor progression and metastasis: is the microenvironment to blame? Cancer Lett 2013; 341: 97–104.

Pattabiraman

Weinberg

. Tackling the cancer stem cells—what challenges do they pose? Nat Rev Drug Discov 2014; 13: 497–512.

10.

Singh

Hawkins

Clarke

. Identification of human brain tumour initiating cells. Nature 2004; 432: 396–401.

11.

Lam

. Side population in human lung cancer cell lines and tumors is enriched with stem-like cancer cells. Cancer Res 2007; 67: 4827–4833.

12.

Chan

. Identification and characterization of tumorigenic liver cancer stem/progenitor cells. Gastroenterology 2007; 132: 2542–2556.

13.

Lee

Zheng

. Cd133+ hcc cancer stem cells confer chemoresistance by preferential expression of the akt/pkb survival pathway. Oncogene 2008; 27: 1749–1758.

14.

Zhang

Shi

Yen

. A subpopulation of cd133(+) cancer stem-like cells characterized in human oral squamous cell carcinoma confer resistance to chemotherapy. Cancer Lett 2010; 289: 151–160.

15.

Shmelkov

Jun

St Clair

. Alternative promoters regulate transcription of the gene that encodes stem cell surface protein ac133. Blood 2004; 103: 2055–2061.

16.

Boivin

Labbe

Fontaine

. The stem cell marker cd133 (prominin-1) is phosphorylated on cytoplasmic tyrosine-828 and tyrosine-852 by src and fyn tyrosine kinases. Biochemistry 2009; 48: 3998–4007.

17.

Corbeil

Roper

Hellwig

. The human ac133 hematopoietic stem cell antigen is also expressed in epithelial cells and targeted to plasma membrane protrusions. J Biol Chem 2000; 275: 5512–5520.

18.

Hilbe

Dirnhofer

Oberwasserlechner

. Cd133 positive endothelial progenitor cells contribute to the tumour vasculature in non-small cell lung cancer. J Clin Pathol 2004; 57: 965–969.

19.

Hemmati

Nakano

Lazareff

. Cancerous stem cells can arise from pediatric brain tumors. Proc Natl Acad Sci USA 2003; 100: 15178–15183.

20.

Chu

Clanton

Snipas

. Characterization of a subpopulation of colon cancer cells with stem cell-like properties. Int J Cancer 2009; 124: 1312–1321.

21.

Chiou

Huang

. Positive correlations of Oct-4 and Nanog in oral cancer stem-like cells and high-grade oral squamous cell carcinoma. Clin Cancer Res 2008; 14: 4085–4095.

22.

Yang

Gupta

. Pentaspan membrane glycoprotein, prominin-1, is involved in glucose metabolism and cytoskeleton alteration. Biochemistry (Mosc) 2007; 72: 854–862.

23.

Griguer

Oliva

Gobin

. Cd133 is a marker of bioenergetic stress in human glioma. PLoS One 2008; 3: e3655.

24.

Moon

Kim

Sohn

. Effect of cd133 overexpression on the epithelial-to-mesenchymal transition in oral cancer cell lines. Clin Exp Metastasis 2016; 33: 487–496.

25.

Toden

Tran

Tovar-Camargo

. Epigallocatechin-3-gallate targets cancer stem-like cells and enhances 5-fluorouracil chemosensitivity in colorectal cancer. Oncotarget 2016; 7: 16158–16171.

26.

Gelbard

Garnett

Abrams

. Combination chemotherapy and radiation of human squamous cell carcinoma of the head and neck augments ctl-mediated lysis. Clin Cancer Res 2006; 12: 1897–1905.

27.

De Felice

Musio

Tombolini

. Head and neck cancer: metronomic chemotherapy. BMC Cancer 2015; 15: 677.

28.

. Targeting cancer stem cells: a new therapy to cure cancer patients. Am J Cancer Res 2012; 2: 340–356.

29.

Westphal

Dewson

Czabotar

. Molecular biology of bax and bak activation and action. Biochim Biophys Acta 1813; 2011: 521–531.

30.

Jin

Chen

. Involvement of microRNA-141-3p in 5-fluorouracil and oxaliplatin chemo-resistance in esophageal cancer cells via regulation of PTEN. Mol Cell Biochem 2016; 422: 161–170.

31.

Munakata

Uemura

Tanaka

. Cancer stem-like properties in colorectal cancer cells with low proteasome activity. Clin Cancer Res 2016; 22: 5277–5286.

32.

Nogales

Reinhold

Varma

. Epigenetic inactivation of the putative DNA/RNA helicase slfn11 in human cancer confers resistance to platinum drugs. Oncotarget 2016; 7: 3084–3097.

33.

Herman

Chen

Garnett

. Comparison of carboplatin-paclitaxel to docetaxel-cisplatin-5-flurouracil induction chemotherapy followed by concurrent chemoradiation for locally advanced head and neck cancer. Oral Oncol 2014; 50: 52–58.

34.

Kim

Jin

. Upregulated expression of BCL2, MCM7, and CCNE1 indicate cisplatin-resistance in the set of two human bladder cancer cell lines: t24 cisplatin sensitive and t24r2 cisplatin resistant bladder cancer cell lines. Investig Clin Urol 2016; 57: 63–72.

35.

Noordhuis

Holwerda

Van der Wilt

. 5-fluorouracil incorporation into RNA and DNA in relation to thymidylate synthase inhibition of human colorectal cancers. Ann Oncol 2004; 15: 1025–1032.

36.

Torok

Paal

. [Chromatographic determination of novamidazofen in pharmaceutical preparations]. Acta Pharm Hung 1997; 67: 193–199.

37.

Galluzzi

Senovilla

Vitale

. Molecular mechanisms of cisplatin resistance. Oncogene 2012; 31: 1869–1883.

38.

Shen

Pouliot

Hall

. Cisplatin resistance: a cellular self-defense mechanism resulting from multiple epigenetic and genetic changes. Pharmacol Rev 2012; 64: 706–721.

Ectopic overexpression of CD133 in HNSCC makes it resistant to commonly used chemotherapeutics